← Directory

ChromaDB

Vector database for AI applications with persistent embeddings storage.

  • AI
  • Databases
  • Tools & Utilities
Vector database for AI applications with persistent embeddings storage. Compatible with Odysseus AI workspace.

Platform availability

Available on 2 of 9 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

84/100 Excellent

Maintained
100last commit 8 days ago
Popular
8929,460 GitHub stars
Easy to install
452 install platforms, install notes, Docker image
Light to run
No data

Based on 3 of 4 factors.

Availability: platform listings confirmed today, links checked 15 days ago

GitHub stars
29,460
Open issues
916
Last commit
2026-10-06
Latest release
1.5.9
License
Apache-2.0
Activity
steady

Checked 8 days ago - source: GitHub (chroma-core/chroma)

Resources & Compatibility

ARM (e.g. Raspberry Pi) is supported.

Checked 17 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Start ChromaDB and note the data path Install ChromaDB and start it with `chroma run --path ./getting-started`, or use the Docker command with `-p 8000:8000` if you prefer containers. The `--path` is where your data is stored, and port 8000 must be reachable by other apps or they won't connect. source
  • Use the provided .env with Compose For the Docker Compose setup, run `docker compose --env-file ./.env up --build` instead of a plain `up`. The `.env` file supplies required settings for the multi-tenant/basic-auth example, and `--build` rebuilds the image so your current code is used. source
  • Save the AdminClient data folder If you create databases with Python's AdminClient (the object used to manage ChromaDB databases), set `is_persistent=True` and point `persist_directory` to a folder you can back up. That folder stores database and tenant information, so losing it can break or make it hard to recover your multi-user setup. source

Fetched 17 days ago - each note links to its source

Alternatives (4)

  • Algolia - Algolia helps product teams connect their users with information by providing the building blocks they need to create fast, relevant, perso…
  • Elasticsearch - Elasticsearch is a distributed search and analytics engine that indexes and queries full-text, log, metric, and vector data in near real ti…
  • Manticore search - Open source search server designed to be fast, scalable and with powerful and accurate full-text search capabilities derived from Sphinx se…
  • OpenSearch - OpenSearch is a scalable, flexible, and extensible open-source software suite for search, analytics, and observability applications derived…

Checked 14 days ago - source: AlternativeTo

Common questions

Can ChromaDB run inside a Python app without a separate server?

Yes. It can be embedded in your Python application and does not need a client-server model.

Where is ChromaDB data stored in client-server mode, and what port must be reachable?

Use the --path flag when running chroma run to set the storage directory. If other apps connect to it, port 8000 must be reachable, and Docker can map it with -p 8000:8000.

How should a multi-user ChromaDB setup created with Python's AdminClient be backed up?

Create the AdminClient with is_persistent=True and point persist_directory to a folder you can back up. That folder stores database and tenant information, so losing it can make recovery difficult.

How often are tagged ChromaDB packages released?

New tagged versions of the pypi and npm packages are released on Mondays. Hotfixes can go out at any time during the week.

Answers sourced from cookbook.chromadb.dev, cookbook.chromadb.dev, github.com, www.reddit.com

Community

Collected 16 days ago - public community threads